Growth Composition and Carbon Intensity in Saudi Arabia: A Regime-Sensitive Predictive Analysis

Climate-change mitigation in resource-dependent economies requires lowering carbon intensity (CO2 per unit of GDP) while sustaining growth. Using annual data for Saudi Arabia (1970–2022), we examine how growth composition—structural change, educational scale, and urban population scale—relates to carbon intensity, via quantile regression and Toda–Yamamoto Granger (predictive) causality. All relationships are short-run, estimated on first-differenced, stationary data without cointegration, with bootstrap inference (standard errors, confidence intervals, pseudo-R2). Energy intensity is the dominant predictor of carbon intensity (≈+1.0 at every quantile, p < 0.001), consistent with the Kaya identity. Among the composition variables, only urban population scale shows a robust, regime-dependent association, turning positive and significant at the 10% level at the upper-middle quantiles (τ = 0.7–0.8). Structural change and trade openness show no robust independent association once energy intensity is controlled; educational scale is insignificant throughout; and no structural break appears around 2014, 2016, or 2020. The findings support a regime-sensitive, diagnostic reading of Vision 2030 prioritizing energy-mix decarbonization and cautious attention to urban demographic scale over uniform composition instruments. The study informs Sustainable Development Goals 7, 8, 9, 11, and 13.